occurrence. On the upper side of the leaf the presence of the pits is denoted by well-marked protuberances. The structure of the leaf and tissues round the pit (fig. 1) is as follows: In transverse section on the upper surface of the leaf there is first a cuticle (cu.) with an irregular outline; beneath this a layer of small epidermal cells (ep.); next a hypodermal layer of colourless, roundish cells, elongated in a transverse direction, and functioning as a storage layer for water (st. l.); beneath this again a layer of cells (ho.) approximating to, the storage layer, but containing in parts a few chlorophyll granules, though in other parts they are quite empty. This layer is longer than broad, thus approaching in shape the palisade cells (p. p.) of the next few rows. These palisade cells, which are typically elongated in a longitudinal direction, contain a great number of chlorophyll corpuscles (chl.), crowded rather closely together towards the centre. They also contain, in almost all the species examined, drops (d.) of what appeared to be oil. Beneath the cells of the palisade parenchyma are those of the spongy parenchyma (sp.) of irregularly shaped cells, with numerous airspaces between. They also contain numerous chlorophyll corpuscles (chl.), though not so many nor so closely crowded as those of the palisade parenchyma. In the spongy parenchyma, too, in all the species I examined I found numerous bundles of crystals or raphides (r.). Beneath the spongy tissue, which occupied the greater part of the leaf-area, is the epidermis of the lower surface (l. ep.), in which are numerous stomata (st.). On the lower surface there is also a cuticle (cu.) with the same irregular outline as that of the upper surface. The stomata are enclosed by two guard-cells (g.), in transverse section smaller and different in shape from the ordinary-epidermal cells. The epidermal cells next to the guard-cells are modified to form subsidiary cells (s.), into which the guard-cells fit, and which overhang the guard-cells. The pit is circular in outline, with projecting rims. Its epidermis (p. ep.) is continuous with that of the lower surface, and similar to it. Above the pit epidermis are two layers of roundish cells (c. l.), slightly elongated in a transverse direction, and colourless in contrast with the spongy tissue immediately surrounding the innermost layer. A few of the epidermal cells are occasionally prolonged into papillose projections or hairs (h), which are usually unicellular, though I saw one such hair consisting of two cells. These hairlike projections are not very numerous, and many pits appeared to be without them. Dr. A. N. Lundstrom regards these pits as hairless. There are no stomata in the pits either of this species or any of the other species I examined. The vascular bundle (v. b.) of the midrib is seen beside the pit, with xylem

above and phloem beneath. It has the usual structure, and is accompanied by parenchymatous cells (par.) and the leaf-sheath (l. s.). Above the xylem are a few round chlorophyll-containing cells (ch.), with few chlorophyll corpuscles (chl.). They are inserted between the xylem and the parenchyma cells of the upper part of the leaf. In sections that have been stained it is seen that the hairs and epidermis of the pit and the colourless layers inside the epidermis have all living protoplasmic contents. This was the case in every species examined, in which all or any of these cells were present.

Fig. 3 is a surface section of the pit showing the stomata in surface view, and the position of the pit between the midrib (m.) and the primary veins. The epidermis of the pit (p. ep.) is seen, the hairs (h.), and the colourless layers (c. l.) inside the epidermis. The characters already mentioned are thus repeated in this section, though the shape of the cells differs in correspondence to the different direction in which the section was cut.

Coprosma lucida. Forster, Char. Gen., 138.

The pits (p.) are situated in the axils of the midrib (m.) and primary veins, and occur with great regularity. I saw none in the axils of the secondary veins. The two leaves figured (figs. 5 and 11) are both varieties of this species. The structure of the tissues as seen in transverse section (fig. 6) is as follows: On the upper surface of the leaf a thick cuticle (cu.) with a wavy outline; an epidermal layer (ep.) of narrow cells elongated in a transverse direction; then a hypodermal layer of roundish cells (st. l.), also elongated in a transverse direction, and serving as a storage layer for water; then two or three layers of typical palisade cells (p.p.), with chlorophyll corpuscles (chl.) and drops of oil (d.). The chlorophyll corpuscles in this case are less crowded, though still numerous. Next comes typical spongy parenchyma (sp.) of irregularly shaped cells, with chlorophyll corpuscles scattered in them. This, as usual in all the species, occupied the greatest thickness of the leaf. Below this is the epidermis of the lower surface (l. ep.), the cells of which are smaller and broader than those of the upper; then a cuticle, thinner, but with the wavy outline of that on the upper surface. In the lower epidermis are numerous stomata (st.) with small guard-cells (g.), and subsidiary cells (s.), not so unlike ordinary epidermal cells. The cuticle forms a projection (p.) at the entrance of the stoma, thus narrowing it and diminishing the rate of transpiration. The epidermis (p. ep.) and cuticle (cu.) of the pit are continuous with and similar to those of the lower surface. There are numerous multicellular hairs (h) in the pit or at its mouth. The individual cells of these

hairs are almost as broad as long; their number varies. Fig. 9 shows one of these hairs stained with eosine. The nuclei (n.) and protoplasmic contents are then clearly seen. In the young pit the hairs are unicellular, as in fig. 10. Next to the epidermis are two layers of colourless round cells (c. l.), next to which again is the ordinary spongy parenchyma of the leaf. The chlorophyll cells that are present in the projecting rims of the pit have fewer corpuscles, and are sometimes almost colourless. This was the case in all the species I examined. All the pits, too, had the same projecting rims and circular outline.

Coprosma propinqua. A. Cunningham, Precurs., n. 472.

Fig. 22 shows the under-surface of a leaf of this species, which, however, is rather broader-leaved than usual. The pits are few in number; sometimes one, sometimes two, and occasionally none are present. The pits (p.) are in the axils of the midrib (m.) and primary veins. The veins also are few in number, as visible to the naked eye. The structure of the leaf in transverse section (fig. 8) is as follows: On the upper surface a rather thick cuticle (cu.) with minute irregular projections; an epidermis (ep.) of very large cells, almost square in shape, which themselves can function as a storage layer for water; then two or three layers of palisade cells with the usual contents, the chlorophyll corpuscles being irregularly scattered. Then comes the spongy tissue of the leaf, of the usual typical form, with slightly less numerous corpuscles than the palisade cells; below that the epidermis of the lower surface (l. ep.), of much smaller cells than that of the upper, and with a thinner cuticle (cu.), which, however, has the same irregular outline. In the lower epidermis are numerous stomata (st.), which have much the same form as those of Coprosma lucida, only the cuticle does not project so much in front of the opening. The guard-cells (g.) are pear-shaped, and the epidermal cells (s.) next to them are slightly modified to fit into them. The epidermis of the pit (p. ep.) is like that of the lower surface, though, of course, there are no stomata; the cuticle (cu.) has a very irregular outline. In the pit are hairs (h.) of two, three, or sometimes more cells. The individual cells of the hairs in this case are long and narrow.

Coprosma linearifolia. Hook. f., Handbk., 118

The upper and lower surfaces of the leaf are seen in fig. 23. The upper surface shows no venation to the naked eye; the lower surface only the midrib and one or two primary veins. The pit (p.) on the under-surface has a tiny opening; on the upper a marked protuberance (pro.). The pits ap-

peared mostly near the end of the leaf, in the axil of the-midrib, and veins not visible to the naked eye. They are usually two in number, placed opposite to one another; but sometimes there are no pits at all present. The structure of the leaf in transverse section is seen in fig. 14. The leaf is-very similar to that of Coprosma propinqua. On the upper surface a cuticle with a surface slightly irregular; an epidermis of large cells, almost square in shape, that may function as a storage layer (ep.); two or three layers of typical palisade parenchyma (p. p), with the usual contents, the chlorophyll corpuscles (chl.) being irregularly scattered; then typical spongy parenchyma (sp.), with less numerous though-still many corpuscles, also irregularly scattered; then the epidermis of the lower surface (l. ep.) of much smaller cells; a cuticle (cu.) with a wavy outline; and numerous stomata (st.), with pear-shaped guard-cells (g.), and slightly modified subsidiary cells (s.). There are, however, no hairs present in the pit. Next to the epidermis of the pit (p. ep.), which is similar to that of the lower surface, and has a cuticle (cu.) with the same wavy outline, is a single layer of roundish-colourless cells (c. l.); next to that a layer (x.) of cells approximating to these in shape, but containing a few chlorophyll corpuscles. These cells are intermediate between the colourless layer and the spongy parenchyma which surrounds-them.

Coprosma fœtidissima. Forst., Char. Gen., 138.

The upper and lower surfaces of the leaf are shown in fig. 24. The pits are few in number, and are situated in the axils of the midrib (m.) and primary veins. On the lower surface no veins are visible to the naked eye, and the opening of the pit (p.) appears to be very small. On the upper surface a slight protuberance (pro.) is seen, as well as the primary veins (p. v.). The structure of the leaf in transverse section is as follows (fig. 15): On the upper surface a cuticle (cu.) with a wavy outline; an epidermis (ep.) of cells slightly elongated in a transverse direction; beneath that a layer of cells (ho.) intermediate between a hypodermal storage layer and the palisade cells. They are like the palisade cells in shape, but contain very few chlorophyll corpuscles. Beneath them are two or three layers of palisade cells (p. p.), with numerous chlorophyll corpuscles (chl.) massed together in the middle of the cells as almost globular aggregates. These cells are broader than are those of the species hitherto described. Below them comes typical spongy parenchyma, with less numerous chlorophyll corpuscles irregularly scattered; and on the lower surface is an epidermis (l. ep.) of small cells, with a thin cuticle (cu.) and numerous stomata (st.). The

guard-cells (g.) are small and pear-shaped; the subsidiary cells (s.) are slightly modified to fit into them. The cuticle does not show the wavy outline of the upper surface. The pit has numerous unicellular hairs (h.) inside and at the opening; they are outgrowths of epidermal cells. Next to the pit epidermis are two layers of roundish cells (c. l.), slightly elongated in a transverse direction, and colourless. Next to these is a layer (x.) of cells, intermediate in shape between the colourless layers and the cells of the spongy parenchyma, and containing very few chlorophyll corpuscles. Above this again is the spongy tissue of the leaf.

Coprosma chathamica. Cockayne, Trans. N.Z. Inst., xxxiv., p. 317.

The under-surface of the leaf is shown in fig. 13. The pits (p.) occur regularly in the axils of the midrib and primary veins. The venation is reticulate. The upper surface of the leaf has only a slight protuberance marking the presence of the pits. The structure of the leaf in transverse section (fig. 16) is as follows: On the upper surface a rather thick cuticle (cu.) with slightly wavy outline; an epidermal layer (ep.) of cells elongated in a transverse direction; a storage layer (st. l.) of roundish cells, also elongated in a transverse direction; two or three layers of typical palisade tissue (p. p.), with chlorophyll corpuscles arranged regularly around the walls of the cells. Beneath the palisade tissue is typical spongy parenchyma (sp.), the cells of which are less irregular in shape than are those of all the species already described. They contain chlorophyll corpuscles, also arranged, though with less regularity, along the walls of the cells. On the lower surface is an epidermis (l. ep.) of smaller cells, with a thinner cuticle. The epidermis of both the upper and lower surfaces has numerous outgrowths or hairs (h.), usually unicellular. There are numerous stomata (st.) on the lower surface, which do not differ from those already described. The epidermis of the pit (p. ep.), which is similar to the epidermis of the lower surface, is a single layer of colourless cells (c. l.), round in shape; next to that is a layer (x.) intermediate in shape between the colourless layer and the cells of the spongy parenchyma, and containing very few chlorophyll corpuscles. The hairs (h.) of the pit are long and multicellular; the individual cells of the hairs are narrow and elongated.

Coprosma petiolata (Chatham Island variety). Hook. f., in Journ. Linn. Soc., Bot., i. (1857), 128.

The under-surface of the leaf is seen in fig. 12. The pits (p.) are arranged with considerable regularity in the axils of

the midrib (m.) and primary veins. The venation is reticulate, the leaf thin and hairy, and the pits (p.), with tufts of hairs projecting from the opening, easily visible to the naked eye. The structure of the leaf in transverse section is very similar to that of Coprosma chathamica. It is seen in fig. 17. On the upper surface a cuticle (cu.) with slightly irregular outline; an epidermis (ep.) of cells elongated in a transverse direction; two or three layers of typical palisade cells (p. p.), with chlorophyll corpuscles (chl.) arranged regularly along the walls of the cells; below that spongy tissue, the cells of which are rounded, and scarcely irregular at all in shape. They contain chlorophyll corpuscles arranged along the walls of the cells. On the lower surface is an epidermis (l. ep.) of smaller cells, with a thin cuticle. The layer of cells next to the epidermis contains very few chlorophyll corpuscles. On the lower surface are numerous stomata (st.) with pear-shaped guard-cells (g.), and slightly modified subsidiary cells (s.). There are many hairs (h.) at the mouth of the pit, which project in tufts. They are multicellular, and some are very long; the individual cells of the hairs are elongated and narrow. The epidermis (p. ep.) and cuticle are similar to those of the lower surface. Next to the pit epidermis are two layers of colourless cells (c. l.), roundish in shape, and slightly elongated in a transverse direction. In this case the layer (x.) of Coprosma chathamica has been fully modified to form a second colourless layer. Around these colourless cells is the spongy tissue of the leaf. There are hairs (h.) also within the pit, but they are not so numerous nor so long as those at the opening. On the upper and lower surfaces of the leaf also there are numerous hairs, mostly two-celled, though they may consist of more cells, or be unicellular. They are especially numerous proceeding from the epidermis next to the parenchymatous cells (par.) surrounding a vascular bundle. Both of these Chatham Island species have less projecting rims to their pits.

Coprosma robusta. Raoul., in Ann. Sc. Nat., ii. (1844), 121.

The under-surface of the leaf is shown in fig. 19. The venation is reticulate. The pits (p.) are regularly arranged in the axils of the midrib (m.) and primary veins. On the upper surface their presence is indicated by a well-marked protuberance. The structure in transverse section (fig. 18) is as follows: On the upper surface a rather thick cuticle (cu.) with slightly irregular outline; an epidermis (ep.) of roundish cells, large and elongated in a transverse direction, and probably serving to store up water in addition to their other functions. Beneath the epidermis are two or three layers of typical palisade cells with the usual contents, the chloro-

phyll corpuscles (chl.) being rather closely crowded, especially towards the centre; beneath the palisade parenchyma is typical spongy parenchyma (sp.), with numerous chlorophyll corpuscles (chl.) scattered irregularly. On the lower surface is an epidermis of small roundish cells (l. ep.), with a thinner cuticle than that of the upper, and numerous stomata. The guard-cells (g.) project slightly, are small and pear-shaped; the subsidiary cells (s.) are slightly modified epidermal cells. The epidermis of the pit (p. ep.) is similar to that of the lower surface, and has a cuticle (cu.). There are numerous projecting hairs (h.) consisting of two or three narrow elongated cells. Next to the epidermis are two layers of round colourless cells (c. l.), and around these is the spongy tissue of the leaf.

Coprosma cunninghamii. Hook. f., Handbk., 113.

The under-surface of the leaf is seen in fig. 20. The pits (p.) are arranged regularly in the axils of the midrib (m.) and primary veins. The structure of the leaf in transverse section (fig. 25) is as follows: On the upper surface a cuticle (cu.) with slightly irregular outline; next an epidermis (ep.) of rather large cells elongated in a transverse direction; beneath these two or three layers of typical palisade cells with the usual contents, and the chlorophyll corpuscles (chl.) rather closely crowded; beneath this, again, typical spongy parenchyma (sp.), with less crowded though still numerous chlorophyll corpuscles; then the epidermis of the lower surface (l. ep.), consisting of small cells, and with a cuticle (cu.) and numerous stomata (st.). The guard-cells (g.) are pear-shaped as usual, the subsidiary cells (s.) slightly modified to fit into them. There are, numerous hairs (h.) at the opening of the pit, as well as inside. They are multicellular, the individual cells of each being narrow and elongated. The epidermis of the pit (p. ep.) is similar to that of the lower surface. In immediate contact with it is the spongy tissue of the leaf, no colourless layers being present inside the epidermis.

Coprosma rotundifolia. A. Cunn., Precurs., n. 472.

The under-surface of the leaf is seen in fig. 21. The pits are few in number, arranged in the axils of the midrib (m.) and primary veins. The leaf is thin. The structure in transverse section (fig. 26) is as follows: On the upper surface of the leaf is a thin cuticle (cu.) and numerous hairs composed of two or three narrow elongated cells; an epidermal layer (ep.) of large cells, elongated in a transverse direction, and probably able to function as a storage layer for water. Beneath this there are two or three layers of palisade cells (p.p.), broader than usual, and with their numerous chlorophyll

garded these pits as coming under the head of “domatia,” and discussed four interpretations that might be put upon them: (1) They may be pathological; (2) they may have only an indirect connection with their tenants; (3) they may be for catching insects; (4) they may be of use to the plant as the dwellings of commensals. He adopts this last interpretation regarding the pits as being the result of a symbiosis between the plant and minute Acarids, which are supposed to make them their homes. He appears to have always found Acarids in possession of these cavities. Mr. Hamilton says that as often as not he found no mites there, and in two instances in which he found large numbers of them the pits were damaged by their presence. In these damaged pits he found “brownish patches here and there, and also in places a number of the cells were of a bright-crimson colour.” I have avoided calling these pits “domatia,” for they do not seem to me to come within the definition of such as given by Dr. Lundstrom—i.e., “those formations or transformations in plants adapted to the habitation of guests, whether animal or vegetable, which are of service to the plant.”

In all the species of Coprosma I examined I found not a single trace of an Acarid in the pits. As some of these species were obtained from the Christchurch Botanical Gardens, or from private gardens, I thought that this might be accounted for by the fact of their not growing naturally in the bush; but in specimens obtained from the bush at Dunedin I found the same state of affairs. In several of the species the tissues round the pits were in an unhealthy condition, with brownish patches here and there, and some cells of a bright-crimson colour, just as described by Mr. Hamilton. There were, however, no Acarids present in such pits. I invariably found in these cases what appeared to be the hyphæ of some fungus growing between the cells of the leaf-tissue. The crimson cells were round in outline, and looked very much like spores, and some seemed to be in the act of germinating. In the cavity of the pits I frequently found quantities of dust or other foreign matter collected, as did also Dr. Lundstrom. The unhealthy state of the pits seemed to me to be due to the attacks of some fungal parasite. The cells of the leaf between which the hyphæ grew were not at all swollen or distorted. The unhealthy condition, too, was by no means confined to the tissue around the pits, but occurred in all parts of the leaf, so that it could scarcely have been due to the presence of Acarids in the pits, though, of course, the injurious effect of such presence would tend to overthrow Dr. Lundstrom's theory as to their utility to the plant. As these mites appear to take refuge in any suitable cavity (Mr. Hamilton found them in rolled leaves of Ricino-

carpus in the stomatal crypts of Banksia, and other cracks and cavities suitable for shelter), their occasional presence in the pits may be accounted for by the fact of these pits presenting a suitable place for shelter. Therefore I incline to think that there is no connection between the formation of the pits and the visits of the insects.

Mr. Hamilton also in his paper discusses and dismisses the probability of the pits being pathological in nature; being glands; being cavities for sheltering stomata; or being extra growths caused by the superabundance of sap. As these suppositions have been sufficiently discussed, I will merely notice in passing the following facts in reference to them: There was no sign of any diseased condition of the leaf-tissues; except in the case already mentioned; there was no sign of any secretion from the hairs or epidermal cells of the pits; there were no stomata in any of the pits. Mr. Hamilton mentions, finally, the possibility of these pits being organs for the absorption of water or vapour, and dismisses this theory also owing to his inability to fill the pits with water. I tried first to fill the pits with an alcoholic stain, as did Mr. Hamilton, and succeeded without any difficulty. The epidermal cells and hairs of the pit in a few minutes showed a faint tinge of the colour of the staining material in their cells. But, lest the alcohol should affect the absorbing-capacity of the living cells in any way, I tried covering the surface of the leaf with water in which was mixed finely powdered carmine. After leaving it for a short time on the surface of the leaf I washed off all traces of the carmine, so that in the action of cutting I might not introduce any of it into the pits. I then cut sections through the pits, and found that there were numerous grains of carmine adhering to the cells of the epidermis and hairs of the pit. The hairs may serve especially to suck up water, though the epidermis probably aids in the process, for in Coprosma linearifolia there are no hairs in the pits, and in some other species they are not numerous enough to be of themselves of great importance. The presence of a cuticle outside the epidermal cells and hairs need be no hindrance to the theory of the pits as absorbing organs, for it is not often thick. A cuticle, unless very thick, is not impermeable to water; and even when very thick it is not so much its thickness that prevents the absorption of water through it, but the presence of and amount of wax in the cuticle, and the contents of the cells it covers. As to the contents of the cells, I can say nothing for certain. I could see in the hairs, in the cells of the colourless layers, and the epidermis of the pit, minute or larger round bodies, some of which may have been drops. They were clear and transparent looking. I could also see these in the cells of the hypo-